[gmx-users] Re: My problem about the job under bond constrain?

[Anton Feenstra]

Hanming wrote:
> I did the steep EM with all-bonds constrained,get LINCS WARNING.
> MD without constrains at a timestep of 2fs is OK, but MD with constrains at
> 2fs get LINCS WARNING.
> MD without constrains at a timestep of 3 or 4fs get a lot of nan in the
> md.log and confout.gro files.
> The following is my grompp.mdp file. When use constrains What I changed  is
> just change the last term constrains = none to constrains= all-bonds. Where
> is the problem?

Probably in the conformation/coordinates of your system. Check it carefully
for bad contacts etc. If you do em, mdrun will tell you which atoms get the
highest forces, those are the first you should check.

A general equilibration setup that works good for me & my type of systems:
1) minimize solute in vacuum, no constraints
2) minimize solute in vacuum, with constraints
3) solvate
4) minimize solute & solvent, no constraints
5) minimize solute & solvent, with constraints
6) MD solute & solvent with position restraints on solute (optionally with constraints)
7) 'production' run with 'free' MD (with constraints & no position restraints)

-- 
Groetjes,

Anton
  _____________ _______________________________________________________
|             |                                                       |
|  _   _  ___,| K. Anton Feenstra                                     |
| / \ / \'| | | Dept. of Pharmacochem. - Vrije Universiteit Amsterdam |
|(   |   )| | | De Boelelaan 1083 - 1081 HV Amsterdam - Netherlands   |
| \_/ \_/ | | | Tel: +31 20 44 47608 - Fax: +31 20 44 47610           |
|             | Feenstra at few.vu.nl - www.few.vu.nl/~feenstra/         |
|             | "If You See Me Getting High, Knock Me Down" (RHCP)    |
|_____________|_______________________________________________________|